pilot-watchdog
Security monitoring for suspicious network patterns in Pilot Protocol networks. Use this skill when: 1. You need real-time detection of suspicious connection patterns 2. You want automated alerts for security anomalies 3. You need to monitor trust relationship changes continuously Do NOT use this skill when: - You only need audit logs (use pilot-audit-log) - You're doing one-time security checks (use pilot-verify)
소스 정보
- 저장소
- TeoSlayer/pilot-skills
- 최근 소스 활동
- 2026년 5월 19일 03:16
- 감지된 SKILL.md 언어
- 영어
- 스타
- 8
- 포크
- 2
설치 방법
기본적으로 소스를 먼저 확인하는 Prompt가 선택됩니다. 직접 명령으로 전환하거나 로컬 사본을 다운로드할 수도 있습니다.
소스 파일 검토
설치 여부를 결정하기 전에 SKILL.md와 SkillsMP에 표시된 보조 파일을 읽어 보세요.
파일 탐색기
2 개 파일SKILL.md 표시 중
SKILL.md
소스 지침 · 읽기 전용 미리보기- name
- pilot-watchdog
- description
- Security monitoring for suspicious network patterns in Pilot Protocol networks. Use this skill when: 1. You need real-time detection of suspicious connection patterns 2. You want automated alerts for security anomalies 3. You need to monitor trust relationship changes continuously Do NOT use this skill when: - You only need audit logs (use pilot-audit-log) - You're doing one-time security checks (use pilot-verify)
- tags
- ["pilot-protocol","trust-security","monitoring"]
- license
- AGPL-3.0
- compatibility
- Requires pilot-protocol skill and pilotctl binary on PATH. The daemon must be running (pilotctl daemon start).
- metadata
- {"author":"vulture-labs","version":"1.0","openclaw":{"requires":{"bins":"[Truncated]"},"homepage":"https://pilotprotocol.network"}}
- allowed-tools
- ["Bash"]
# Pilot Watchdog
Real-time security monitoring for Pilot Protocol with anomaly detection, automated alerting, and threat response.
## Commands
### Initialize Watchdog
```bash
mkdir -p ~/.pilot/watchdog/{alerts,state}
cat > ~/.pilot/watchdog/config.json <<EOF
{
"enabled": true,
"check_interval_seconds": 30,
"rules": {
"connection_rate_limit": 10,
"failed_handshake_threshold": 3,
"queue_drop_threshold": 1
}
}
EOF
```
### Monitor Connection Rate
```bash
# Detect abnormal connection rate
CURRENT=$(pilotctl --json connections 2>/dev/null | jq -r '.[].remote_hostname' | sort | uniq -c)
echo "$CURRENT" | while read -r COUNT AGENT; do
if [ "$COUNT" -gt 10 ]; then
echo "ALERT: $AGENT has $COUNT connections"
fi
done
```
### Monitor Queue-Drop Counter
The daemon's `health` payload exposes `queue_drops` — when this is non-zero, the accept queue is overflowing (CPU saturation, fd limit too low, or DoS). Detect a regression by snapshotting and diffing.
```bash
STATE_FILE=~/.pilot/watchdog/state/queue_drops.txt
CURRENT=$(pilotctl --json health | jq -r '.queue_drops // 0')
PREVIOUS=$(cat "$STATE_FILE" 2>/dev/null || echo 0)
if [ "$CURRENT" -gt "$PREVIOUS" ]; then
DELTA=$((CURRENT - PREVIOUS))
echo "ALERT: queue_drops increased by $DELTA since last check (now $CURRENT)"
fi
echo "$CURRENT" > "$STATE_FILE"
```
### Set Webhook for Events
```bash
# Configure daemon webhook
pilotctl --json set-webhook "http://localhost:8080/watchdog"
```
## Workflow Example
Continuous security monitoring with automated responses:
```bash
#!/bin/bash
WATCHDOG_DIR=~/.pilot/watchdog
INTERVAL=30
mkdir -p "$WATCHDOG_DIR"/{alerts,state}
# Alert function
alert() {
echo "$(date -u +%Y-%m-%dT%H:%M:%SZ) [$1] $2" | tee -a "$WATCHDOG_DIR/alerts/alerts.log"
}
# Monitor loop
while true; do
# Check connection rate
pilotctl --json connections | jq -r '.[].remote_hostname' | sort | uniq -c | \
awk '$1 > 10 {print $2}' | while read agent; do
alert "CONNECTION_RATE" "$agent exceeded connection limit"
done
# Check failed handshakes
PENDING=$(pilotctl --json pending | jq length)
if [ "$PENDING" -gt 5 ]; then
alert "HANDSHAKE" "High number of pending handshakes: $PENDING"
fi
sleep $INTERVAL
done
```
## Dependencies
Requires `pilot-protocol` skill, `pilotctl` binary, `jq`, and running daemon.
GitHub에서 보기